Abstract Sarcomas primarily affect children and young adults. For pediatric oncologists, sarcomas are a major problem. Most are extremely aggressive and current protocols of surgery, radiation and chemotherapy are not adequate. Significant improvement for sarcoma treatment will require a detailed understanding of the genetic and epigenetic factors responsible for their progression. Each sarcoma is likely to need its own specific molecular-based therapies. To achieve these goals, will likely require generating a mouse model of the human cancer that accurately simulates the human condition in order to increase the information base for each sarcoma. The criterion to establish a ?good? mouse model for human cancers is now very rigorous. Each human cancer is classified by a genetic fingerprint comprised of the expression profile of thousands of genes. Having established authenticity of the mouse model, it can be used to identify the genetic events responsible for tumor progression. Having in turn identified the genetic players, selected components can be used as targets to generate a more rational approach to therapeutic development. We have successfully modeled alveolar rabdomyosarcoma, synovial sarcoma, clear cell sarcoma and alveolar soft-parts sarcoma. We will use selected mouse models to determine the secondary genetic events responsible for sarcoma progression. The contribution of the identified secondary events to malignancy will be genetically assessed in cell lines derived from the respective tumors as well as in the mouse models themselves. We also propose to generate improved third generation mouse models of sarcomas that initiate cancer by inducing the appropriate chromosomal translocation in their cell of origin, rather than utilizing conditional expression of the fusion gene in selected tissues. These third generation mouse models should provide a significant improvement over current conditional expression mouse models by much more closely recapitulating the cancer etiology observed in humans.